feat: handle malformed xml file

pipeline/beam_etl/helpers.py

@@ -1,14 +1,15 @@
+import logging
 import xml.etree.ElementTree as ET
+
 from typing import Dict
 
+
 def iter_parse(root: ET.Element) -> Dict[str, str]:
-    """Recursively parse the XML tree into a dictionary
+    """Recursively parse the XML tree into a dictionary Each key/value pair is
-    Each key/value pair is inside it's own <div> tag and
+    inside its own <div> tag and the key inside a <b> tag.
-    the key inside a <b> tag.
+    The fields that I believe are compulsory (TCIN, UPC and Origin) are only
-    The fields that I believe are compulsory (TCIN, UPC and Origin)
+    nested one level deep, while the rest of fields seem to be always nested
-    are only nested one level deep, while the rest of fields seem
+    two levels deep. But parsing it recursively helps generalise both cases."""
-    to be always nested two levels deep. But parsing it recursively
-    helps generalise both cases."""
 
     spec_dict = {}
     for child in root:
@@ -22,10 +23,11 @@ def iter_parse(root: ET.Element) -> Dict[str,str]:
                 spec_dict.update(iter_parse(child))
     return spec_dict
 
+
 def parse_raw_specs(raw_specs: str) -> Dict[str, str]:
-    """Parse a raw specifications XML string into a dictionary
+    """Parse a raw specifications XML string into a dictionary.
-    This involves first recursively parsing the XML tree and then
+    This involves first recursively parsing the XML tree and then renaming
-    renaming the key values"""
+    the key values"""
 
     fields_mapping = {
         "Material": "materials",
@@ -37,7 +39,13 @@ def parse_raw_specs(raw_specs: str) -> Dict[str,str]:
         "TCIN": "tcin",
         "Origin": "origin",
     }
+
+    try:
         xml_root = ET.fromstring(raw_specs)
+    except ET.ParseError:
+        logging.error("error parsing xml string: \n%s", raw_specs)
+        return {}
+
     parsed = iter_parse(xml_root)
     specs_dict = {
         fields_mapping[key]: value